Effects of Slum Demolitions: Evidence from Victorian London

A growing majority of the world’s population lives in cities, many of which suffer from a proliferation of slums and the resulting extreme inequality in living conditions. Slums are poor environments to raise families, yet may also offer residents employment, informal support networks, and the chance to improve their lot. As cities develop, slums are often demolished to make room for better housing or local amenities. However, despite the ubiquity of slum demolitions, we know surprisingly little about their impacts. Our study fills this gap by empirically examining the long-run impacts of slum demolitions, compared to a counterfactual of leaving the slums in place. We do so using slum demolitions in late 19th-century London as a natural experiment. London slums were sites of poverty, bad living conditions, and high death rates, resembling in many respects the slums in developing countries today. However, starting in the late Victorian era, slums were demolished as part of public efforts to remove poor-quality residences and make way for better housing. We digitize the addresses of buildings that were slated to be demolished as part of 59 slum demolition projects. We then link the addresses to the pre-demolition residents, drawing on historical full-count population censuses that contain both name and address data. Finally, we use automated linking procedures to follow slum dwellers and their children over time. Crucially, only 30 of these 59 demolition projects were actually carried out. This allows us to use the residents of undemolished slums as counterfactual for the residents of demolished slums. We then estimate the effects of slum demolition both on the individual slum residents and on whole areas of the city. We find that demolished slum addresses were 55 percentage points less likely to be inhabited 10 years later compared to undemolished ones. Initial slum residents got displaced and were 9 percentage points less likely to remain at their old residences. At the area level, slum demolitions were associated with gentrification, as the share of white-collar workers in a demolished neighbourhood increased by 10 percentage points. However, this gentrification did not benefit the initial slum residents. They tended to move to areas with similar socioeconomic characteristics as their initial slum residences, and we further find evidence that their social networks got disrupted. Yet, despite these and other stressors, the disruptions from having been displaced do not seem to have worsened the labour market outcomes of affected slum residents.